Incidence and prognostic significance of non-enhancing cortical signal abnormality in glioblastoma

Introduction The presence of non-enhancing cortical signal abnormality is useful for differentiating between glioblastoma and metastatic disease, but its significance has not been studied. We aimed to determine the incidence and prognostic implications of non-enhancing cortical signal abnormality in glioblastomas. Methods Patients with a new diagnosis of glioblastoma between September 2007 and December 2010 were identified. Only patients with at least fluid-attenuated inversion recovery (FLAIR) and post-contrast T1-weighted sequences were included. Pre-operative MRIs were reviewed together by two readers to determine whether there was evidence of non-enhancing cortical signal abnormality, based primarily on the FLAIR sequence. The results were compared with patient survival using both univariate and multivariate analysis. Results One hundred fifty-one patients met the inclusion criteria. Seventy-seven patients (51%) had evidence of non-enhancing cortical signal abnormality. On both univariate and multivariate analysis, there was overall no significant difference in survival between patients with non-enhancing cortical signal abnormality and those without. Peripheral enhancing lesions (51 patients) were generally associated with the longest survival, but subgroup analysis suggested that non-enhancing signal abnormality was associated with worse survival in these patients (P = 0.004), conveying an intermediate prognosis. Conclusions Non-enhancing cortical signal abnormality is a common feature of glioblastomas, occurring in about half of cases. It does not affect survival overall, but appears to be associated with worse survival in the setting of a peripherally located enhancing lesion. This has the potential to alter surgical management.